| 00:04 | <phroggy> | also, in my case, I want the little magnifying glass icon to indicate the purpose of the box, because I don't want to take the space to add a label, and using a placeholder isn't appropriate because I automatically focus on page load (so the user can immediately start typing to filter search results). |
| 01:50 | <phroggy> | there's something I've never understood: why is target="_blank" deprecated? The official workaround is to simulate this behavior with JavaScript, which is obviously appropriate in some situations but in others just seems like reinventing the wheel. |
| 01:50 | <Hixie> | aha, the video of my talk is available internally now. should be public soon. |
| 01:51 | <Hixie> | i say "um" far too much. |
| 01:51 | <Hixie> | wonder how to work on that. |
| 01:51 | <Hixie> | phroggy: target=_blank isn't deprecated in html5 |
| 01:51 | <phroggy> | it isn't? oh good. |
| 01:51 | <phroggy> | it was in 4.01, right? |
| 01:52 | <Hixie> | target="" wasn't in HTML4 strict at all |
| 01:52 | <phroggy> | well, right... |
| 01:52 | <phroggy> | so, in 5, is it back in, or still left out? |
| 01:52 | <Hixie> | html5 has target="" |
| 01:52 | <phroggy> | "deprecated" was the wrong word, I suppose. |
| 01:52 | <Hixie> | in general though i strongly recommend against using target="" at all, creating new windows for the user is not a good experience |
| 01:52 | <phroggy> | it can be. |
| 01:53 | <Hixie> | should be the user's choice imho |
| 01:53 | <Hixie> | i basically just stop using sites that think they know better than me when to open windows |
| 01:54 | <phroggy> | for example, I'm making a contact directory where I ask the user for various things like their MySpace/Facebook/LiveJournal names, or the URL to their home page or whatever, and on the confirmation screen I offer these as URLs they can check. I believe it would be totally inappropriate for these to open in the same window/tab. |
| 01:56 | <phroggy> | it should be the user's choice whether to open them in a new window or a new tab, but navigating away from the current page is not desirable. Obviously I use target="_blank" very sparingly, but simulating it with JavaScript (as was recommended by W3C for 4.01 Strict) is just ridiculous. |
| 02:09 | <Hixie> | if i was on that page, i would expect the links to open on the same page and to use the back button to return to the form |
| 02:09 | <Hixie> | and it would piss me off if a new tab or window opened |
| 02:09 | <Hixie> | to the point of not using the site again |
| 02:13 | <phroggy> | except that this is the result of a POST form, and some browsers don't like to cache those. |
| 02:15 | <phroggy> | I believe that most of my users would find your preferred behavior disruptive. Obviously I could be mistaken. |
| 02:32 | <phroggy> | Hixie: what's your opinion of <input type="search">? |
| 02:32 | <othermaciej> | Hixie: well, we're publicly claiming full pass of Acid3 now, I hope enough time has passed that the test won't change again |
| 02:32 | <othermaciej> | also I don't know if you or anyone else intends to officially verify |
| 06:17 | <BenMillard> | phroggy, I would also expect those links to be normal links, but I'd probably open each one in a new background tab since I'd only need to see the page title to confirm it's the right page. |
| 06:41 | <karlUshi> | annevk2: here? |
| 06:41 | <karlUshi> | I have a simple question about html5lib |
| 06:42 | <karlUshi> | Does the serializer drop the elements which are not authorized by the HTML 5 content model? |
| 06:42 | <karlUshi> | such as center for example |
| 06:52 | <Hixie> | phroggy: i like type=search |
| 06:58 | <karlUshi> | hmm interesting |
| 06:58 | <karlUshi> | http://code.google.com/p/html5lib/source/browse/trunk/?r=947#trunk/python/src/html5lib/serializer%3Fstate%3Dclosed |
| 06:59 | <karlUshi> | I'm looking at the code of the serializer and it doesn't seem to clean up elements which are not included in the content model |
| 06:59 | <karlUshi> | which would make them non conformant HTML 5 "writer" |
| 06:59 | <karlUshi> | s/them/it/ |
| 07:01 | <zcorpan_> | karlUshi: that's not the responsibility of a serializer, though, is it? |
| 07:02 | <karlUshi> | zcorpan_: well it depends what you mean by serializer. |
| 07:02 | <Hixie> | BenMillard: is there any documentation on exactly what the smart headers algorithm is and exactly how it differs from the "smart headers"-based algorithm in the spec |
| 07:02 | <karlUshi> | if the role of the serializer is to output an *HTML 5 document* |
| 07:03 | <karlUshi> | I would say that it should remove the obsolete elements. no? |
| 07:03 | <karlUshi> | if the serializer is a garbage in/garbage out, indeed it doesn't emit html 5 markup |
| 07:03 | <zcorpan_> | right |
| 07:04 | <karlUshi> | I wonder if someone has worked on a product taking the parsed document and outputing conformant html5 |
| 07:04 | <zcorpan_> | you could have a layer before the serializer to strip out stuff |
| 07:04 | <zcorpan_> | or convert it to html5 equivalent |
| 07:05 | <karlUshi> | hixie: is it in the plan to define the rules for sanitizing the output? |
| 07:05 | <zcorpan_> | karlUshi: in my mind a serializer is just the reverse of a parser: emit a stream of characters (or bytes) from a tree |
| 07:06 | <zcorpan_> | karlUshi: everyone would want different rules for sanitizing |
| 07:06 | <karlUshi> | zcorpan_: yes I can accept this, but it should clearly be mark as such in the spec. I haven't checked yet though :) |
| 07:07 | <karlUshi> | zcorpan_: hmm about different rules. Well, I think that would benefit many people and remove a bit of anger. ;) I'm trying to address this in blog post. |
| 07:08 | <karlUshi> | so I stand corrected for the serializer thanks. |
| 07:08 | <karlUshi> | I will fix the article. (not published yet) |
| 07:08 | <zcorpan_> | ok |
| 07:08 | <hsivonen> | karlUshi: html5lib also has a sanitizer. |
| 07:09 | <hsivonen> | karlUshi: it is for xss security, not conformance, though, afaik |
| 07:09 | <karlUshi> | yep seen that in the ruby version |
| 07:10 | <zcorpan_> | karlUshi: there's http://wiki.whatwg.org/wiki/Sanitization_rules |
| 07:11 | <karlUshi> | it keeps center which is not in html5 |
| 07:11 | <karlUshi> | (for example) |
| 07:11 | <karlUshi> | understood that it is for security |
| 07:12 | <Hixie> | karlUshi: sanitising what output? |
| 07:12 | karlUshi | didn't use sanitizing at the start of the discussion |
| 07:13 | <karlUshi> | just saying |
| 07:13 | <karlUshi> | Garbage In -> HTML 5 Parser -> HTML 5 Writer (output = html 5 conformant document) |
| 07:14 | <Hixie> | given garbage in, there is no way that i know of in the state of the art today to give conformant output |
| 07:14 | <zcorpan_> | print "<!doctype html><title></title>" |
| 07:14 | <karlUshi> | the serializers in HTML 5 lib do not fix the markup to be conformant according to html 5 spec |
| 07:14 | <zcorpan_> | karlUshi: so you want an HTML5Tidy |
| 07:15 | <karlUshi> | yes which would be basically html5 serializer + html5 content model filter |
| 07:15 | <BenMillard> | Hixie, not in complete and absolute detail in one place, AFAIK. |
| 07:15 | <zcorpan_> | karlUshi: it would probably also have to include SVG Tidy and MathML Tidy :) |
| 07:15 | <karlUshi> | to create a virtuous circle |
| 07:16 | <BenMillard> | Hixie, I recall James Graham wrote up a snapshot of what was implemented some time ago but that doesn't compare it to what's in HTML5...ah here it is: http://lists.w3.org/Archives/Public/public-html/2008Mar/0075.html |
| 07:16 | <Hixie> | BenMillard: k |
| 07:16 | <Hixie> | BenMillard: thanks |
| 07:17 | <Hixie> | BenMillard: ah yeah i used that when writing the spec |
| 07:17 | <Hixie> | iirc |
| 07:17 | <Hixie> | BenMillard: (though i made some changes because the spec had other constraints to worry about as well) |
| 07:17 | <BenMillard> | yeah, I think you mentioned it in #whatwg |
| 07:20 | <BenMillard> | Hixie, the main difference I know if is the chaining of header cells. my vision for Smart Headers is it would be implemented by UAs, make all possible associations through the table, pass this on via platform-specific accessibility APIs, then ATs would use their already very sophisticated verbosity controls to only provide the relationships that are useful to their specific customer base (and individuals would have the abilit |
| 07:20 | <BenMillard> | as they can customise verbosity already) |
| 07:21 | <BenMillard> | in the tables I studied during 2007, only 20% used <th> for all their header cells (http://projectcerbera.com/web/study/2007/tables/) |
| 07:22 | <BenMillard> | I didn't count how common <td scope="row"> was for row headers compared to <th> or <th scope="row"> |
| 07:23 | <BenMillard> | but my impression was that <td scope> was common enough to be worth supporting |
| 07:23 | <Hixie> | i need to look at the feedback in the tables directory |
| 07:24 | <Hixie> | but my impression from the data i got yesterday was that chained headers hardly ever happen in the wild |
| 07:24 | <Hixie> | and when they do, almost never happen correctly |
| 07:24 | <Hixie> | and so it's probably not worth adding that feature |
| 07:24 | <Hixie> | i certainly am not convinced we want to chain headers automatically |
| 07:25 | <BenMillard> | I don't mean chaining of the headers attribute, I mean cases where a <th> has another <th> which is related to it, such as a table with 2 or more levels of column headers |
| 07:25 | <Hixie> | but again, i'll have to look at the data and feedback closely when i get to this issue again |
| 07:25 | <Hixie> | the thing i don't understand is why would a user ask for the headers of a header? |
| 07:25 | <Hixie> | surely it's the data the user is looking for |
| 07:25 | <Hixie> | and the data cells already have those headers applied |
| 07:26 | <hsivonen> | creating a comprehensive html5 tidy will be hard |
| 07:26 | <hsivonen> | for example, it would need to fix tables with overlapping cells |
| 07:27 | <karlUshi> | hsivonen: didn't say it would be easy. It has certainly a lot of issues, but a set of canonical rules would improve interop and would help people to move forward. |
| 07:29 | <hsivonen> | much of it could be done by hacking jing internals |
| 07:34 | <hsivonen> | the 'curies are not qnames' explanation is interesting when curies share the characteristic of qnames being criticized |
| 07:35 | hsivonen | had a look at xhtml2 minutes |
| 07:35 | <BenMillard> | Hixie, you sometimes need to understand the context of a header cell as you move through a table non-visually. A table I wrote recently, in the "Comparison of Techniques" section here: http://projectcerbera.com/blog/2008/09/untangle |
| 07:36 | <BenMillard> | if you're moving down the 2nd column, it's probably useful to hear when you've moved out of the "Test File" section into the "Corrected" section, etc |
| 07:37 | <BenMillard> | another one from me, a bit older, at the bottom of this page: http://projectcerbera.com/web/articles/breadcrumbs-markup |
| 07:37 | <BenMillard> | moving across the 2nd row, it's probably useful to hear when you moved out of "Filesize (bytes)" column group and into the "Familiar" column group |
| 07:41 | <zcorpan_> | BenMillard: send what you just wrote in email :) |
| 07:41 | <BenMillard> | zcorpan_, good idea |
| 07:54 | <karlUshi> | many thanks henri, zcorpan_ and hixie. Published. rotten tomatoes as usual in the comment box :) |
| 07:58 | <zcorpan_> | pointer? |
| 07:58 | <Hixie> | BenMillard: yeah, i could see that |
| 07:58 | <karlUshi> | zcorpan_: http://www.w3.org/QA/2008/09/fixing-html-with-html5 |
| 07:59 | <Hixie> | BenMillard: not sure it's really common enough to warrant adding to the spec, but i'll have to look closer |
| 08:01 | <BenMillard> | Hixie, for sure. but it is a regular layout and it does exist outside of my website, for example: http://projectcerbera.com/web/study/2007/tables/odi/06-collated/original |
| 08:03 | <Hixie> | might make sense to just have headers have headers and treat them the same was cells are treated, without doing any nesting |
| 08:03 | <BenMillard> | Hixie, the original source for that is still online, "Table 6" on this page: http://www.odi.govt.nz/publications/going-beyond-compliance/survey/appendix-4-task-speed.html |
| 08:03 | <Hixie> | (that gets around the problem of loops) |
| 08:03 | <Hixie> | (though it doesn't get around the problem of self-reference) |
| 08:03 | <zcorpan_> | karlUshi: choise of quotes is a bad example since you're discussing content model and quotes is syntax :) |
| 08:03 | <karlUshi> | zcorpan_: but I'm pretty the question will arise :) |
| 08:04 | <BenMillard> | Hixie, the original source uses plain <th> and expects it to Just Work (which seems reasonable to me) |
| 08:04 | <karlUshi> | in fact will love to loose a lot of times on these points ;) |
| 08:04 | <karlUshi> | s/fact/fact, people/ |
| 08:04 | <Hixie> | BenMillard: those tables will work fine as far as i can tell |
| 08:05 | <karlUshi> | s/pretty/pretty sure/ |
| 08:05 | karlUshi | in plain demonstration of my article :p unfortunately. Clumsy one day, clumsy for ever |
| 08:08 | <BenMillard> | Hixie, when moving across the 2nd row of "Table 6" it's probably useful to hear when you move from "Fastest" to "Slowest"? |
| 08:08 | <BenMillard> | (similar arrangement as my breadcrumbs-markup table) |
| 08:09 | <BenMillard> | in fact, I'd say it's essential since the header text is the same for the columns within those 2 column groups |
| 08:09 | <BenMillard> | so you need the main column header from row 1 to disambiguate the column headers in row 2 |
| 08:15 | <Hixie> | yeah, i could see that |
| 08:28 | <zcorpan_> | "<Hixie> might make sense to just have headers have headers and treat them the same was cells are treated, without doing any nesting" -- yeah that'd make sense |
| 08:31 | <BenMillard> | annevk2 has a table where hearing headers as you move through headers seems useful: http://lists.w3.org/Archives/Public/www-archive/2007Aug/att-0003/offset-mess.htm |
| 08:31 | <BenMillard> | specifically, moving across row 2 and row 3 |
| 08:32 | <BenMillard> | (I think the top-left cell could be blank and the browser names could use <th>) |
| 08:32 | <annevk2> | sounds reasonable |
| 08:33 | <annevk2> | karlUshi, dunno much about the serializer, but it seems it worked out already |
| 08:34 | karlUshi | thanks annevk2. Will look further. Need to catch my train. |
| 08:34 | <BenMillard> | so multiple levels of column headers are somewhat easy to find...but multiple levels of row headers? maybe I'm the only person who does that :P |
| 08:38 | <BenMillard> | aha, my 2007 collection has this: http://www.avert.org/aofconsent.htm |
| 08:39 | <BenMillard> | "Australia", "United Kingdom" and "USA" in the first column have a 2nd column of row headers giving smaller regions within them |
| 08:40 | <BenMillard> | so if you're moving down the 2nd column of "United Kingdom" it's probably useful to hear when you fall into "USA" |
| 08:49 | annevk2 | wonders what to make of http://wiki.whatwg.org/wiki/Special:Contributions/Stephen_Shea |
| 08:56 | <annevk2> | http://lists.w3.org/Archives/Member/chairs/2008JulSep/0112.html (W3C Member-only) |
| 08:59 | <annevk2> | http://lists.w3.org/Archives/Member/w3c-ac-forum/2008JulSep/0273.html (W3C-Member-only, also interesting) |
| 09:07 | <zcorpan_> | (Testing: see http://krijnhoetmer.nl/irc-logs/.) |
| 09:15 | <annevk2> | https://bugzilla.mozilla.org/show_bug.cgi?id=373864#c4 is also interesting (and public!) |
| 09:17 | <BenMillard> | Hixie, I just summarised in an on-list e-mail the tables I was discussion earlier where header cells being announced when moving between other header cells seemed useful to me: http://lists.w3.org/Archives/Public/public-html/2008Sep/0475.html |
| 09:17 | <Hixie> | cool, thanks |
| 09:17 | <Hixie> | i'll add it to the tables pile |
| 09:17 | <BenMillard> | cheers :) |
| 09:18 | <Hixie> | annevk2: Larry's right, though I think he's missed the WHATWG side of things |
| 09:23 | <othermaciej> | well that's interesting |
| 09:28 | <othermaciej> | annvevk2: I find this one interesting: <http://lists.w3.org/Archives/Member/w3c-ac-forum/2008JulSep/0189.html>; (W3C Member-only again) |
| 09:29 | <othermaciej> | annevk2: in light of the fact that any rational look at working groups and their effectiveness would surely put XForms and XHTML2 on the cut list |
| 09:29 | <Hixie> | given a=1234 and b=5678, what is the mathematical expression that yields 1234.5678 ? |
| 09:29 | <annevk2> | othermaciej :) |
| 09:29 | <othermaciej> | Hixie: well, the simplest such expression would be 1234.5678 |
| 09:30 | <othermaciej> | if you want it to involve a and b, then a + b / 1000 |
| 09:30 | <othermaciej> | if a and b are actually general variables with varying numbers of digits |
| 09:31 | <Hixie> | i mean given any value of a and b, with the result being equivalent to casting a and b to a string and then concerting a + "." + b to an integer again, except without going through strings. |
| 09:31 | <othermaciej> | then a + b * (10 ^ -(1 + floor(log10(b)))) |
| 09:31 | <othermaciej> | or something like that |
| 09:32 | <Hixie> | pity |
| 09:32 | <BenMillard> | do you get a bonus for connecting 4 closing parentheses in a row? :) |
| 09:32 | <othermaciej> | former lisp hacking habits |
| 09:33 | <Philip`> | f(x) = { x if x < 1; f(x/10) otherwise }; a+f(b) |
| 09:33 | <Philip`> | That's still a mathematical expression :-) |
| 09:33 | <othermaciej> | if one were writing this in code then yes a loop dividing by 10 might be more effective |
| 09:34 | <othermaciej> | of course, this assumes a and b are integers |
| 09:34 | <othermaciej> | though if they are not the question is ill-posed |
| 09:35 | annevk2 | wonders what problem this is solving |
| 09:35 | <Philip`> | Or you could write "a.b" and people may know what that means |
| 09:36 | <Philip`> | (I'm sure I've seen that notation several times in the past) |
| 09:53 | annevk2 | is a bit surprised that the smart header algorithm only now gets attention, especially after yesterday's telcon where some people were stressing for deciding on this right "now" |
| 09:56 | <Hixie> | annevk2: really? you're surprised? after the way they've been acting? |
| 10:03 | <annevk2> | maybe disappointed is better |
| 10:04 | <Hixie> | i wish someone would tell me what the rush was about, frankly |
| 10:04 | <Hixie> | but anyway |
| 10:04 | <Hixie> | time for bed |
| 10:04 | <Hixie> | nn |
| 10:05 | <othermaciej> | Hixie why do you hate blind people so much |
| 10:24 | <takkaria> | oh, hsivonen is porting his java parser to Gecko |
| 10:31 | <annevk2> | yeah, he said as much anyway |
| 10:36 | <hsivonen> | yes, I'm porting it |
| 10:37 | <annevk2> | lot of work? |
| 10:37 | <hsivonen> | the parser core is suitable for mechanic translation to C++ |
| 10:37 | <hsivonen> | we'll see :-) |
| 10:38 | <annevk2> | might be annoying to keep the two up to date all the time I imagine |
| 10:38 | <olliej> | annevk2: prod? |
| 10:38 | <hsivonen> | I expect the integration to Gecko to be more work than the Java to C++ translator |
| 10:39 | <annevk2> | but maybe someone else can do maintenance after you did the hard work? |
| 10:39 | <annevk2> | olliej, yo |
| 10:39 | <hsivonen> | annevk2: the idea is to keep them in sync by generating C++ code from the Java source |
| 10:39 | <olliej> | annevk2: xhr spec says that the XHR object's document is the document of the object it is invoked on |
| 10:39 | <olliej> | annevk2: which works if 'this' is a window |
| 10:40 | <othermaciej> | "object it is invoked on" is kind of an ambiguous phrase |
| 10:40 | <olliej> | annevk2: but what happens if i do (for the sake of argument) new ({a:XMLHttpRequest}.a) |
| 10:40 | <olliej> | annevk2: what document should it get? |
| 10:40 | <othermaciej> | I told olliej I thought it meant the "this object" |
| 10:41 | <olliej> | annevk2: esp. if you imagine windowa.foo = function(){new ({a:windowb.XMLHttpRequest}.a) } |
| 10:42 | <annevk2> | sigh, I'm not sure what that syntax does |
| 10:42 | <othermaciej> | annevk2: olliej is setting it up so that the object that the XMLHttpRequest constructor is invoked on, is not a Window object |
| 10:42 | <othermaciej> | more readable example: |
| 10:42 | <othermaciej> | var randomObj = {}; |
| 10:42 | <othermaciej> | randomObj.XMLHttpRequest = someWindow.XMLHttpRequest; |
| 10:43 | <othermaciej> | new randomObj.XMLHttpRequest(); |
| 10:43 | <annevk2> | hmm, I never tried that, does that work in IE? |
| 10:44 | <othermaciej> | I bet IE binds the original window object the XHR constructor comes from |
| 10:44 | <othermaciej> | rather than using the one it is invoked on |
| 10:44 | <othermaciej> | you could tell the difference with: |
| 10:45 | <othermaciej> | window.foreignXHR = otherWindow.XMLHttpRequest |
| 10:45 | <annevk2> | what the spec says is what IE does for the non randomObj cases |
| 10:45 | <othermaciej> | new foreignXHR(); |
| 10:45 | <othermaciej> | what does IE do in that case? |
| 10:45 | <othermaciej> | does it use window or otherWindow? |
| 10:45 | <annevk2> | i'm not sure about that case either, I don't have IE handy |
| 10:46 | <othermaciej> | I would guess then that it was not the intent of the spec to voice an opinion on these cases |
| 10:46 | <othermaciej> | olliej: I guess you should test in IE and maybe Firefox and Safari and Opera and send your results to the webapps-wg mailing list |
| 10:49 | <annevk2> | http://tc.labs.opera.com/apis/XMLHttpRequest/open/ has tests for the "simple" client = new otherwindow.XMLHttpRequest() cases |
| 10:49 | <annevk2> | it does seem like this needs addressing, indeed |
| 11:28 | annevk2 | finds http://lists.w3.org/Archives/Member/w3c-wai-pf/2008JulSep/0509.html (W3C-Member only) |
| 11:35 | <othermaciej> | mmm, condescenscion |
| 11:42 | <annevk2> | for the window.foreignXHR = otherWindow.XMLHttpRequest case I'd expect window.document to be the Document pointer |
| 11:42 | <Dashiva> | It's good to know nothing goes on in the w3c that needs hiding |
| 11:43 | <annevk2> | the PFWG should really be public, not sure why it isn't |
| 11:43 | <othermaciej> | they are the experts |
| 11:43 | <othermaciej> | why should they need outside input? |
| 11:44 | <othermaciej> | annevk2: if you expect that, you expect the "this" object to be used, but in that case the randomObject case should fail |
| 11:44 | <Dashiva> | Oh, I would never suggest such a thing, merely that we might want insight in their wisdom |
| 11:44 | <Philip`> | They should open up and try to emulate the great success of public-html |
| 11:44 | <annevk2> | othermaciej, I feel another HTML5 dependency coming |
| 11:45 | <othermaciej> | public-webapps works pretty well |
| 11:46 | <othermaciej> | I think public-html fails because the launch of the HTML WG got so much publicity |
| 11:46 | <othermaciej> | and because HTML is such a central piece of the Web technology stack |
| 11:46 | <othermaciej> | and because it is easier to wrongly believe you understand HTML than to do so with CSS or DOM APIs |
| 11:46 | <hsivonen> | webapps attracts implementors and people who write relatively advanced JS |
| 11:47 | <othermaciej> | these factors contribute to a much higher level of highly opinionated but poorly informed people |
| 11:47 | <annevk2> | www-style has that too for less technical topics |
| 11:48 | <othermaciej> | and people who are there not to provide expertise but to score political points for their pet objectives |
| 11:48 | <othermaciej> | public-webapps is one of the best standards mailing lists I am on |
| 11:48 | <annevk2> | othermaciej, weinig, http://krijnhoetmer.nl/irc-logs/webapps/20080926 |
| 11:48 | annevk2 | hilited the relevant bits |
| 11:49 | <othermaciej> | annevk2: yeah, that's what Gecko does |
| 11:50 | <othermaciej> | which does not appear to match the spec as I read it |
| 11:50 | <othermaciej> | not sure what IE does |
| 11:50 | <othermaciej> | the Gecko behavior would not be hard to implement |
| 11:51 | <annevk2> | ah yes, "The associated Document object is the one returned by the document attribute from the object on which the XMLHttpRequest() constructor was invoked (a Window object)." needs some tweaking |
| 11:57 | <othermaciej> | annevk2: it was actually olliej who pointed it out, not Sam |
| 11:57 | <othermaciej> | olliej == Oliver Hunt; weinig == Sam Weinig |
| 12:17 | <annevk2> | oops |
| 12:18 | <olliej> | annevk2: :D |
| 12:21 | <othermaciej> | olliej: do you have test cases for the rebinding examples discussed a little while ago? |
| 12:21 | <othermaciej> | so someone can test in IE |
| 12:21 | <olliej> | othermaciej: errr, not really |
| 12:22 | <olliej> | othermaciej: problem being that i don't know xhr well enough to make useful tests |
| 12:22 | <olliej> | s/well enough/at all |
| 12:27 | <othermaciej> | ok I might try at some point |
| 12:29 | <annevk2> | I was confused I guess because weinig does XHR2 |
| 12:31 | <othermaciej> | olliej just hacks whatever he feels like |
| 12:31 | <othermaciej> | either that or canvas and XHR have some secret relationship |
| 12:32 | <othermaciej> | "WebKit erfüllt alle Acid3-Kriterien" |
| 12:32 | <othermaciej> | somehow it sounds cooler in German |
| 12:35 | <olliej> | dun dun duhh |
| 14:43 | <annevk2> | http://ajaxian.com/archives/maintaining-css ouch |
| 15:36 | <hendry> | anyone have tips for running IE6 in XP (alongside 8beta2 I currently have running)? |
| 15:40 | <Philip`> | Use the VirtualPC image? |
| 17:01 | <BenMillard> | annevk2, zomg: "</body><body>" |
| 17:03 | <BenMillard> | and the 4 </div> means up to 3 are invalid (although that shouldn't cause real problems, it is ugly) |
| 17:05 | <BenMillard> | oh not it doesn't...Firefox 2 wasn't displaying the end of the conditional comments even when I scrolled right to the end (had to select the sample for them to show up) |
| 17:07 | <BenMillard> | actually, I was right the first time: up to 3 are invalid since there's only 1 <div> start tag in browsers which aren't IE7, IE6 or IE5 |
| 17:10 | <BenMillard> | for some reason I thought Ajaxian was actually good...guess my judgement was clouded by memories of Galaxian :) |
| 19:10 | <tusho> | Is <time>2006-09-23</time> fine? |
| 19:11 | <tusho> | As opposed to <time datetime="2006-09-23">2006-09-23</time> |
| 19:13 | <Dashiva> | As far as I can see, yes |
| 19:13 | <Dashiva> | "If the datetime attribute is not present, then the date or time must be specified in the content of the element, such that parsing the element's textContent according to the rules for parsing date or time strings in content successfully extracts a date or time." |
| 19:34 | <tusho> | *nod* |
| 21:50 | <Hixie> | othermaciej: whatwg doesn't have any of the public-html problems, despite it having far more subscribers |
| 21:51 | <gsnedders> | Hixie: So is it inversely proportional for groups > 1 person? |
| 21:52 | <Hixie> | it's all to do with whether the group has someone actually maintaining the community and actually calling people (privately) on their rudeness |
| 21:53 | gsnedders | sighs deeply |
| 21:57 | <Hixie> | http://www.youtube.com/watch?v=xIxDJof7xxQ |
| 21:59 | <dglazkov> | hey, Hixie: http://codereview.chromium.org/4097 -- I know it's not 100, but better than before :) |
| 22:01 | <Philip`> | It's not 100, it's 500 |
| 22:01 | <Philip`> | Oh, the server error went away when I reloaded |
| 22:02 | <Hixie> | dglazkov: :-) |
| 22:04 | <dglazkov> | Philip`: yeah, we occasionally score 500/100 on acid3. It's rare, but it happens ;) |
| 22:05 | <Hixie> | http://www.w3.org/Style/2008/css-charter -- "Effective participation is expected to consume one work day per week for each participant; two days per week for editors." |
| 22:05 | <Hixie> | hahha |
| 22:19 | <Hixie> | http://www.w3.org/mid/48DCD23D.3040202⊙ac |
| 22:19 | <Hixie> | The 2022 date is working in our favour |
| 22:19 | <Hixie> | because the xhtml2wg is buying it |
| 22:19 | <Hixie> | and thus don't think we're a threat ("years from adoption") |
| 22:20 | <Dashiva> | Unlike xhtml2 which is decades away? ;) |
| 22:23 | <jgraham> | "years from adoptions" could be a simple sign error |
| 22:24 | <jgraham> | Since "adoption staarted years ago" is the correct answer |
| 22:24 | <Dashiva> | What discussion about removing iframes on whatwg@ seems a bit silly |
| 22:25 | <Hixie> | jgraham: hah |
| 22:29 | <gsnedders> | Hixie: in the intro to the talk you use the future tense a lot |
| 22:29 | <Hixie> | d'oh! |
| 22:30 | <Dashiva> | gsnedders: He wrote the talk back in 1999, didn't you know? |
| 22:30 | <gsnedders> | Dashiva: :P |
| 22:32 | <gsnedders> | Hixie: I do like the aim of demoing everyting |
| 22:34 | <gsnedders> | Hixie: Where's the full demo video you use? |
| 23:02 | <Hixie> | gsnedders: http://www.whatwg.org/demos/2008-sept/video/ |
| 23:11 | <othermaciej> | Hixie: whatwg never got the kind of giant publicity explosion in the same way |
| 23:11 | <othermaciej> | Hixie: a least not at laucnh |
| 23:11 | <othermaciej> | Hixie: and it is not conducive to politically oriented types |
| 23:12 | <Hixie> | that was intentional -- i told the w3c not to make a big deal out of it, but they felt compelled to do it anyway |
| 23:12 | <othermaciej> | both due to lack offficial status, and due to the way ou run things |
| 23:12 | <Hixie> | still can't believe they released a _press release_ with the FPWD |
| 23:12 | <Hixie> | so stupid |
| 23:13 | <Hixie> | it's more evidence of how they are trying to keep the w3c alive at the cost of the web, instead of the other way around |
| 23:22 | <othermaciej> | well, they need money and attention can be helpful towards such goals |
| 23:22 | <othermaciej> | still, if the W3C wanted good general PR, there are much better things they could do or be doing |
| 23:22 | <othermaciej> | but they do not seem to have the management in place to do sane things |
| 23:27 | <roc> | hostile takeover by the WHATWG |
| 23:27 | <Hixie> | were it only possible |
| 23:35 | <othermaciej> | I must admit, learning that the W3C fails even at making money shocked me |
| 23:36 | <othermaciej> | I thought it was the one thing they were halfway good at |
| 23:37 | <roc> | they could of course save a lot of money by shutting down all the bogus groups |
| 23:37 | <roc> | but I guess that would also impact revenue |
| 23:39 | <jcranmer> | how does the W3C even get revenue? |
| 23:39 | <Dashiva> | Member fees? |
| 23:39 | <Dashiva> | *Membership |
| 23:39 | <jcranmer> | oh |
| 23:41 | <othermaciej> | is anyone a member just to participate in the winfest that is OWL? |
| 23:42 | <Dashiva> | I figure anyone involved with OWL has interests in all the other SW stuff |
| 23:42 | <othermaciej> | all of which is equally a winfest |
| 23:45 | <roc> | hmm. I assume "winfest" expands to "festival of wins", but that doesn't sound right |
| 23:46 | <Dashiva> | roc: There may or may be some element of sarcasm involved |
| 23:46 | roc | recalibrates his detector |
| 23:46 | <Dashiva> | omg |
| 23:47 | <Dashiva> | Hixie promotes @alt in the video |
| 23:47 | <Hixie> | lies! |
| 23:47 | <Dashiva> | And I used to think you were cool |